Timeline for How fitting a mixture distribution of noncentral Student t components to a one-dimensional sample in R?
Current License: CC BY-SA 3.0
20 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 9, 2022 at 17:14 | comment | added | Kenric | I addressed a related question to estimating the Student's t distribution. The method described could also be applied to the noncentral Student's t. See stats.stackexchange.com/a/570972/207683 | |
| May 23, 2017 at 12:39 | history | edited | CommunityBot |
replaced http://stackoverflow.com/ with https://stackoverflow.com/
|
|
| Apr 13, 2017 at 12:44 | history | edited | CommunityBot |
replaced http://stats.stackexchange.com/ with https://stats.stackexchange.com/
|
|
| Jun 22, 2015 at 18:22 | comment | added | Mark L. Stone | What about mixed Lognormals? Is any one doing that in Finance? Or mixture of Lognormal and some fatter-tailed distribution? | |
| Jan 14, 2015 at 13:41 | history | tweeted | twitter.com/#!/StackStats/status/555358961571225600 | ||
| Jan 13, 2015 at 0:18 | answer | added | Aleksandr Blekh | timeline score: 1 | |
| Jun 22, 2014 at 16:21 | answer | added | Joz | timeline score: 0 | |
| May 2, 2014 at 16:45 | comment | added | Joz | To be precise: I want to set up a dynamic benchmark scenario to evaluate certain methods to estimate VaR and ES with. The mixture will NOT be used to estimate VaR or ES or to model the volatility. But anyway, I`d like to fit a mixture to data in order to calibrate my scenarios to history so they will be sort of realistic. In case this is not possible I will fit individual noncentral scaled student t distributions to periods that reflect certain market regimes. These periods might be chosen with help of theory (looking backward we know the crisis period) and/or with breakpoint tests. | |
| May 2, 2014 at 16:36 | history | edited | Joz | CC BY-SA 3.0 |
added 123 characters in body
|
| May 2, 2014 at 16:28 | comment | added | whuber♦ | It depends on which aspects of risk are important for you to capture. Now that I know you are assessing risk, I appreciate the value of using components that might have infinite moments. | |
| May 2, 2014 at 15:38 | comment | added | Joz | A gaussian mixture would be too unflexible (I don't want to work with more than 2 components at a time due to intuition issues in setting up scenarios). Other distributions easily get even more complicated I think (e.g. skewed distributions). One very important aspect for me is that there should be (implicit) analytical formulas for the Value at Risk and the Expected Shortfall (which I have set up for the t mixture already). So this gives nice opportunities for performance testing of risk measures. But feel free to suggest another alternative. I'm open minded! | |
| May 2, 2014 at 14:51 | comment | added | whuber♦ | What I am asking is why: Why should a noncentral t distribution be a good choice to model "returns" (presumably for some investment prospect)? There are potentially many other more tractable families of qualitatively similar distributions that can do the same thing. What is it about these investments that suggests using noncentral t distributions to the exclusion of all else? | |
| May 2, 2014 at 8:20 | history | edited | Joz | CC BY-SA 3.0 |
added 19 characters in body
|
| May 2, 2014 at 7:37 | comment | added | Joz | I want to use the non-central student t components because I aim at a scenario analysis framework to test the performance of dynamically modeled risk measures such as Value at Risk or Expected Shortfall for a project at uni. A dynamically adjusted mixture distribution of two such components can generally reflect the characteristics of returns. Moreover, the two components can be nicely interpreted as two scenarios (quiet market versus stress period) and therefore subjectively chosen or manipulated. But to fit them to historical data I need some estimation technique for their parameters. | |
| May 2, 2014 at 7:25 | history | edited | Joz | CC BY-SA 3.0 |
added 343 characters in body
|
| May 1, 2014 at 22:42 | comment | added | whuber♦ | I am somewhat curious why you would choose non-central Student-t components, because they are not very computationally tractable: they have only a limited number of moments of small order and most of their properties are difficult to compute. What theory of the genesis of your data leads to such a mixture model? | |
| May 1, 2014 at 22:20 | history | edited | Joz | CC BY-SA 3.0 |
added 101 characters in body
|
| May 1, 2014 at 21:00 | history | edited | Joz | CC BY-SA 3.0 |
added 467 characters in body
|
| May 1, 2014 at 19:36 | review | First posts | |||
| May 1, 2014 at 19:38 | |||||
| May 1, 2014 at 19:19 | history | asked | Joz | CC BY-SA 3.0 |